The sites and mechanisms of postoperative insulin resistance

In Sweden with nine million inhabitants, 450,000 operations (outpatients excluded) are performed every year resulting in 2,250,000 treatment days in hospital. Surgical operations are part of the treatment for 44% of all patients admitted to hospital care occupying 24% of all hospital beds. The majority of these patients undergo an elective surgical procedure. Therefore, it is important to reduce the side effects of surgery, such as the catabolic response.

Insulin is a key anabolic hormone which regulates not only the metabolism of glucose, but also the metabolism of fat and protein. Insulin resistance is a main feature of the catabolic response to surgery and other trauma. However, the sites and mechanisms of the postoperative insulin resistance remain to be clearly defined. Furthermore, it is not known whether changes in postoperative insulin sensitivity have any impact on patient outcome. Therefore, insulin sensitivity and glucose kinetics ([6,6,2H2]-D-glucose) were determined using hyperinsulinemic, normoglycemic clamps and indirect calorimetry, before and after elective surgery in patients undergoing abdominal surgery (n = 18) or total hip replacement (n = 13).

The patients were undergoing surgery after the traditional overnight fast (n = 17) or in a carbohydrate fed state. The carbohydrate fed state was achieved by infusions of glucose and insulin 3-4 hours before and during surgery (n = 7) or by intake of a carbohydrate drink (400 ml, 50 g carbohydrates) 2-3 hours before surgery (n = 7). Glucose infusion rates required to maintain normoglycemia during clamps (M-value) were reduced after surgery in the overnight fasted patients, indicating the development of postoperative insulin resistance. Endogenous glucose production was moderately increased after surgery. The suppressibility of endogenous glucose production by insulin was preserved postoperatively. Thus, most of the reduction in insulin sensitivity after surgery was due to a defect in glucose disposal. Since postoperatively, a similar reduction in both glucose oxidation and nonoxidative glucose disposal was observed, a defect in glucose transport probably underlies a decrease in insulin sensitivity.

Energy expenditure increased after surgery and fat oxidation rates were less suppressed by insulin infusions. Only moderate changes were found in glucagon and cortisol levels after surgery. To single out the effects of surgery from the effects of the common perioperative treatment with bed rest and hypocaloric nutrition, insulin sensitivity was measured in healthy subjects (n =6), before and after a 24 hour period of hypocaloric nutrition and/or bed rest. Insulin sensitivity was reduced after 24 hours hypocaloric nutrition alone while the same period of immobilization had no effect. One group of patients was treated with infusions of insulin and glucose before and during total hip replacement and compared to controls. In the insulin and glucose treated patients, insulin sensitivity remained unaffected immediately after surgery while insulin sensitivity was reduced in the control patients, undergoing the same operation after an overnight fast. A more convenient way to administer carbohydrates would be as a beverage instead of an intravenous infusion.

To test the possibility of administrating carbohydrates orally before the operation, gastric emptying of an isoosmolar carbohydrate rich drink (400 ml, 12% carbohydrates) was determined using gamma camera technique (99Tcm). Despite increased anxiety preoperatively, gastric emptying of the drink was completed 90 minutes after intake in patients in the morning of surgery. When the carbohydrate drink was given to patients 2-3 hours before elective colorectal surgery, postoperative insulin sensitivity was markedly improved as compared to patients undergoing similar surgical procedures after an overnight fast. Multiple regression analysis showed that 72% of the variability in the relative reduction in insulin sensitivity after elective abdominal surgery could be predicted by the duration of surgery (p =0.0002) and the carbohydrate access preoperatively (p = 0.0005).

Furthermore, the length of hospital stay was related to the degree of postoperative insulin resistance (relative change in M-values postoperatively vs hospital stay in postoperative days, r = -0.60, p= 0.018). In addition, 64% of the variability in hospital stay was predicted by the type of surgery (hip or abdominal) (p = 0.0001), duration of surgery (p = 0.010) and whether the patients were fasted or carbohydrate fed before surgery (p = 0.004). Thus, carbohydrate feeding seems to be a better preparation than overnight fasting before surgery by improving postoperative insulin sensitivity and patient recovery following elective surgery.